def transform_trlist_to_data(trList):
"""
"""
timeSteps, rewards = unzip(trList)
data = {"rewards": list(rewards), "time": timeSteps}
return data
def learn(self, actionRewardTupleList):
"""
Learns from a list of ActionReward Tuples
"""
actions, rewards = utils.unzip( actionRewardTupleList )
self.Q.update( newValueOrList=rewards, indexList=actions )
def test_unzip():
x = [1, 2, 3]
y = [4, 5, 6]
z = zip(x, y)
v, w = utils.unzip(z)
assert v == x, "Unzip 1"
assert w == y, "Unzip 2"
def learn(self, actionRewardTupleList):
"""
Learns from a list of ActionReward Tuples
"""
actions, rewards = utils.unzip( actionRewardTupleList )
self.Q.update( newValueOrList=rewards, indexList=actions )
# TODO Add multiple decays
self.decay_epsilon()
def transform_arlist_to_data(arList):
"""
"""
actions, rewards = unzip(arList)
timeSteps = range(len(actions))
data = {
"actions": list(actions),
"rewards": list(rewards),
"time": timeSteps
}
return data